Research map / Graph learning / GNN architectures
Adaptive Universal Generalized PageRank Graph Neural Network
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Use adaptive learned GPR weights instead of fixed APPNP weights | GPR-GNN · Cora | accuracy (%): 79.51 vs 79.41 | No clear effect |
| Use adaptive learned GPR weights instead of fixed APPNP weights | GPR-GNN · CiteSeer | accuracy (%): 67.63 vs 68.59 | Hurt |
| Use adaptive learned GPR weights instead of fixed APPNP weights | GPR-GNN · PubMed | accuracy (%): 85.07 vs 85.02 | No clear effect |
| Use adaptive learned GPR weights instead of SAGE | GPR-GNN · Computers | accuracy (%): 82.90 vs 83.11 | Hurt |
| Use adaptive learned GPR weights instead of APPNP | GPR-GNN · Photo | accuracy (%): 91.93 vs 91.11 | Helped |
| Use adaptive learned GPR weights instead of APPNP | GPR-GNN · Chameleon | accuracy (%): 67.48 vs 51.91 | Helped |
| Use adaptive learned GPR weights instead of APPNP | GPR-GNN · Actor | accuracy (%): 39.30 vs 38.86 | Helped |
| Use adaptive learned GPR weights instead of GCN | GPR-GNN · Squirrel | accuracy (%): 49.93 vs 45.66 | Helped |
| Use adaptive learned GPR weights instead of MLP | GPR-GNN · Texas | accuracy (%): 92.92 vs 92.26 | Helped |
| Use adaptive learned GPR weights instead of APPNP | GPR-GNN · Cornell | accuracy (%): 91.36 vs 91.80 | Hurt |
| Use random initialization for GPR weights instead of informed initializations (ablation) | GPR-GNN · cSBM, dense split | accuracy | Hurt |
| Use random initialization for GPR weights instead of informed initializations (ablation) | GPR-GNN · cSBM, sparse split | accuracy | Hurt |
Closest papers by meaning
- Beyond Fixed Depth: Adaptive Graph Neural Networks for Node Classification Under Varying Homophily (AAAI 2026 · similarity 0.90)
- Demystifying Structural Disparity in Graph Neural Networks: Can One Size Fit All? (NeurIPS 2023 · similarity 0.90)
- Predicting Global Label Relationship Matrix for Graph Neural Networks under Heterophily (NeurIPS 2023 · similarity 0.89)
- Adaptive Node Feature Selection for Graph Neural Networks (ICML 2026 · similarity 0.89)
- Making Classic GNNs Strong Baselines Across Varying Homophily: A Smoothness–Generalization Perspective (NeurIPS 2025 · similarity 0.89)
- On the Impact of Feature Heterophily on Link Prediction with Graph Neural Networks (NeurIPS 2024 · similarity 0.88)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.